persistant cel and drivability issues

[Dusterbd]

will do when it stops raining. hiopefully tomorrow.

[Dusterbd]

The secondary air pump relay and wiring check out. Theres also a considerable amount of air coming from the pump to the valve on the manifold. However, im questioning the electrical switch for the vacuum side. The check valve tesys good, but i dont have vaccum at the valve on the manifold.

Does that help on the diagnostics?

[Dusterbd]

the secondsry sir pump codes are no longer present after replacing the hose between the pump and valve. at least not at this time. additionally I can verify that the valve, line, and check valve all hold vacuum as intended. I still cannot confirm the solenoid as operating though.

current codes, though none show as presently active. I have driven the car for multiple cycles with the CEL lit.

20180912_173219 by Michael Crawford, on Flickr

so it looks as if im back to o2 codes. again.
present from beginning to now is b1s2 no signal. have confirmed all continuity between sensor plug and DME. no shorts or open circuits. therefore it almost HAS to be a faulty sensor. right? is there a way to test with a multimeter to confirm before replacement?

the deviation codes have be stumped. ive smoke tested multiple times with no evidence of vacuum leaks. could it be fuel pressure?

[bmwdirtracer]

Yes, fuel/air mixture errors can be caused by incorrect fuel pressure. You might also have very dirty or leaking injectors, or even a bad MAF.

But (53 posts in, and on my 3rd beer), haven't you swapped plugs between B1S2 and B2S2 already? And ended up watching live data with both sensor readings frozen at .4 ?

When you have done your smoke tests, have you allowed ~10 minutes for the smoke to get all the way to the oil return from the CCV to the dipstick tube? Have you tried removing the oil cap on the valve cover while the engine is running, as a (terrible, unreliable) test of crankcase vacuum?

Sorry, I'm just hunting for possibilities, since you've obviously followed a scientific test regimen....

[Dusterbd]

I have done the b1s2 and b2s2 swap. Did not read live data, or take pictures of the codes. However, the no communication code went away.

Im uncertain if i let the smoke test go for 10 minutes. I have not pulled the oil cap while running. Will do both and report back.

[bmwdirtracer]

Very carefully inspect the oil return hose to the dipstick from the CCV. This hose is usually in a braided sheath, which makes it impossible to see a rotted, melted hose inside, which is a very real major intake system leak. It takes forever for the smoke to get there, even with a pro smoke machine. 10 minutes should do. Opening the oil fill cap until smoke comes out, then replacing it will reduce the time by ~3 minutes.

I'll loan you my slack-tube manometer to test crankcase vacuum, if you like. I'll mail it to you, you can mail it back when you're done. (No charge, of course)

If you feel like driving across the state this hurricane weekend, I'll take home our shop's cheapest ($3500) scan tool, and do some live data readings for you at my home. Really, that's a long damned drive for you, and probably not a viable option when the entire damned population's a;ready driving around avoiding Florence. Still, I live in the western mountains, which will be wet, but that's about all.

[Dusterbd]

I really appreciate the offer. This weekend may not be the best (my wife is convinced that the hurricane is going to wreck the state), but if it continues to baffle me i may take you up on it.

Im going to try smoke testing again, with every variable i can think of eliminated. It either has to be vacuum or fuel delivery.
I picked up a new b1s2 last night to see if that eliminates the no signal code.

[Dusterbd]

Im now CERTAIN there are no vacuum leaks. I turned a wooden plug on my lathe to fit in place of the maf. Propped throttle open. Injected smoke in at nipple behind maf. Got good smoke from open oil fill cap, then closed it. Let machine run for 30 minutes. No wisps of smoke from anywhere, and i even had the car on stands to look underneath from different angles. Built up enough pressure in the system that there was actual force behind the oil fill cap when removed at end of test.

So im certain we can rule out vacuum leaks.

[Dusterbd]

Replaced the b1s2 sensor. No signal errors gone. But still getting b1s1 control deviation too large.

Dirt racer: i think the best use of time and resources may be to buy you a six pack.....

Unless we have other ideas. Cause ive got nothing at this point. Everything seems to check out.

[bmwdirtracer]

OKAY then! PROGRESS!

Now signal errors are gone, we've just got mixture issues, on primary sensors, right? Are you absolutely sure that B1S1 sensor is plugged into B1S1 plug ?

You do have some live data, right? Try unplugging one of the primary sensors, and seeing whether the correct sensor disappears.

Do you have a schrader valve on your fuel rail? It should be right at the front, under the center engine cover.

The E39 has a very common issue with fuel pressure deviations, caused by the combination fuel filter/pressure regulator under the driver's seat area.

[Dusterbd]

Mixture issue only on bank 1 primary sensor.

All are plugged in the correct way, confirmed via multimeter/pinout and visual.

I do have the Schrader valve, however its too small for my fuel pressure gauge to fit.

But would fuel pressure just effect one bank?

[bmwdirtracer]

It's very unlikely that errant pressure would effect just one bank. However, the primary sensor in the other bank is much older, so might not "see" as clearly as the new one. (And let's not forget that the fuel filter is a maintenance item.) I can loan you a schrader valve adapter, by mail, if you want.

Two things come to mind:

An exhaust leak at the header flange or O2 sensor bung (which you'd likely hear).

OR one or more clogged or leaking injectors. It's amazing how bad injectors get, in ~20 years.

I regularly send sets of injectors off to RC Injection, in California, for testing, ultrasonic cleaning, and "restoration", followed by retesting. The printed results have pretty much ALWAYS shown differentials of over 15%, sometimes 20.

They also replace all the o-rings, pintle caps, and make them look better than new, too. $25 per injector.....best bargain in the automotive world.

Of course, you could swap the bank 1 injectors for the bank 2 injectors, clear the codes, and see if the code follows the injectors. Personally, though, I always feel nervous moving and reusing 20 year old rubber o-rings.

[Dusterbd]

They are allowing me to swap the b1s1 under the previous owner warranty. I doubt it will fix anything, but its the only once i haven't replaced yet from the nuked dme.

Nobody has a rental tool that fits the schrader valve.

I replaced all the upper and lower injector o rings when i did the vacuum lines and manifold gaskets, etc. At the same time i saked the injectors and blew through with carb cleaner.

Think a heavy seafoam may be in order to see if it changes things.

[bmwdirtracer]

My friend, sending Seafoam to do an ultrasonic cleaning's job is pissing in the wind, sorry.

Go ahead and swap injectors 1-2-3 with 4-5-6; see if the code follows.

If you PM me with your address, I'll mail you my schrader valve adapter Monday, and you can mail it back after you're done testing. (Last time I checked, I had 2.)

How's your weather? It's just barely beginning to get a bit windy here; rain starts tonight, I'm told.

[Dusterbd]

Its been raining, but nothing worth noting so far. Supposed to be worse this evening.

Swapping the injectors will take a minute, as the rest of the parts to fix the duster showed up today. So after the storm passes, ill haul it to dads place and pull the drivetrain.

When i pulled b1s1 for warrenty replacement, it was sooty. Like a plug thats been run super rich. Which tells me an injector is hanging open. Right?

Will pm address. But its looking like I'll be ordering a set of injectors, i think. Is there a way to bench test to see which one is bad?

[bmwdirtracer]

Do NOT "order a set of injectors". Send yours to RCInjection. The results are documented, and VASTLY cheaper than new injectors. Besides, if you don't have a "before" report, how will you know for sure whether this was the issue?

This is important: If you send your injectors to RC, make absolutely sure to SCRATCH the cylinder number into each injector as you remove them !

Ink or paint won't be there when they get back, and the identification numbers will not match, unless YOU do this scratching.

I await your input on whether your injectors are black in color, following up on your PM.

[Dusterbd]

The injectors are either very dark grey or black. Im going with black, personally.

And sending them out for cleaning is far cheaper than replacement. Wow.

[Dusterbd]

I have three drive cycles without a cel.
No pending codes according to torque. Bmw scanner isn't showing anything either.

Maybe its fixed?

[bmwdirtracer]

Assuming I can actually get to work tomorrow (50 miles away, where it floods a lot), I'll see if I can find a set of injectors to send you, to borrow while you send yours to RC.

- - - Updated - - -

Overlapped your last post.

That's GREAT! When you say "drive cycles", what do you mean? Around the block, or many miles?

[Dusterbd]

First cycle was 30 miles. Shut down 4 hours. 36 miles next, shut down about 5 minutes while i ran into the gas station. 4 miles home after that.

Gonna run back out for a couple of miles here in a couple of hours. See what happens.....

[bmwdirtracer]

I've got my fingers crossed for you, Michael.

[Dusterbd]

Well, definitely progress!

No more mixture codes!

However, flow issues both banks on secondary air injection. Felt almost like a slight miss right before the cel came on.

Is there a good description of the parts of the system and they're operating parameters so i can test individually? I know the air pump runs and puts out a hell 9f a lot of air, and the check valve works.

[shadowpuck]

that's odd though - it's not particularly cool in your area right now, is it? usually, secondary air injection runs at ambient temps at or around 40F at startup if I recall....

i think your version of the secondary air system uses vacuum hose that runs around the back of the engine - there's also a valve or something back there i think.

this is all going off of memory (i know the later system a bit better as that's what is on my car). you can smoke the secondary air system.

[Dusterbd]

Its high 60s here.

Vacuum line around the intake to a valve bolted on the head. Electrical solenoid on the vacuum line. Hose from pump to valve on head.

But how does it all work? Haven't seen a good write up for it yet....

[shadowpuck]

my point was its odd that it is firing up this system when ambient temps are above 40.....
a google search will produce numerous writeups as to how the system works.

the CCA roundel magazine also had a writeup in tech talk awhile back, i think.